Picus Canus) in Boreal Forests

Ornis Fennica 97: 89–100. 2020

Nest tree characteristics of Grey-headed Woodpeckers (Picus canus) in boreal forests

Timo Pakkala*, Juha Tiainen, Heikki Pakkala, Markus Piha & Jari Kouki

T. Pakkala, J. Tiainen, University of Helsinki, Lammi Biological Station, FI-16900 Lammi, Finland. * Corresponding author’s e-mail: [email protected] T. Pakkala, J. Kouki, University of Eastern Finland, School of Forest Sciences, P.O. Box 111, FI-80101 Joensuu, Finland J. Tiainen, Natural Resources Institute Finland, P.O. Box 2, FI-00791 Helsinki, Finland H. Pakkala, Käsityöläiskatu 47 A 2, FI-06100 Porvoo, Finland M. Piha, Finnish Museum of Natural History – LUOMUS, P.O.Box 17, FI-00014 Univer- sity of Helsinki, Finland

Received 20 January 2020, accepted 10 September 2020

Woodpeckers are important species in forest ecosystems because they make tree cavities that are microhabitats for several other taxa. However, even in boreal areas where most tree cavities are made by woodpeckers, the properties of woodpeckers’nest trees and cavities are poorly known. We studied nest tree characteristics of the Grey-headed Wood- pecker (Picus canus) in a 170-km2 forest-dominated area in southern Finland during 1987–2019. The data included 76 nest trees with 80 nest cavities in five different tree species. During the study period, 44% of all nesting attempts were in old cavities. Nests were found in four forest types, but the proportions of nest tree species differed between them. In all, aspen (Populus tremula) with 70% of nest trees, and with 71% of nest cavities was the dominant cavity tree species. Most nest trees (85%) were dead or decaying, and most cavities (70%) were excavated at visible trunk injury spots. The mean diameter of a nest tree at breast height (DBH) was 37.2 cm and the mean height of a cavity hole was 7.8 m; these were significantly positively correlated. The results highlight the importance of large aspens as nest cavity sites for the species. Conservation and retention of groups of large aspens in main habitats, including clear-cuts, are important for continuous availability of nest trees. This applies particularly to managed boreal forest landscapes where scarcity of suitable trees may be a limiting factor for the species.

1. Introduction selves, especially to ensure the excavation of strong and safe cavities, but also for several other Woodpeckers are proposed as indicator or key- cavity-nesting animals like mammals, birds and stone species of forest structural complexity and invertebrates that use these cavities afterwards species diversity (Mikusiñski et al. 2001, Roberge (Jones et al. 1994, Drever et al. 2008, Cockle et al. et al. 2008, Pakkala et al. 2014). The properties of 2011, Siitonen & Jonsson 2012, Hardenbol et al. nest trees are important for woodpeckers them- 2019). Most tree cavities in the boreal zone are 90 ORNIS FENNICA Vol. 97, 2020 suggested to be made by woodpeckers (Aitken & 2.2. Grey-headed Woodpecker Martin 2007, Cockle et al. 2011, Andersson et al. nest tree surveys 2018), and thus are very important microhabitats in these forests. As part of an intensive population study of forest In this study, we investigated nest tree charac- bird species, especially woodpeckers, Grey- teristics of the Grey-headed Woodpecker (Picus headed Woodpecker nests and nest trees were canus) in a boreal forest landscape dominated by searched within the study area each year during the managed coniferous forests. The species has a period 1987–2019. The annual census typically wide Eurasian distribution, and it prefers decidu- lasted from early April to mid-July and included ous tree dominated or mixed forests with edges the mapping of woodpecker territories within the and openings (Dementiev & Gladkov 1966, Glutz study area with simultaneous efforts to locate po- von Blotzheim & Bauer 1980, Cramp 1985, Blu- tential nesting sites by observing the behaviour of me 1996, Saari & Südbeck 1997). Moreover, the the woodpeckers, and by searching for nests dur- Grey-headed Woodpeckers mostly use injured or ing the breeding season (described in detail in decaying deciduous trees for excavating their nest Pakkala 2012 and in Pakkala et al. 2014, 2017). cavities (Conrads & Herrmann 1963, Glutz von Annual territory locations and their approxi- Blotzheim & Bauer 1980, Cramp 1985, Südbeck mate borders were defined by information from 2009), but published information of the nest tree observed locations and movements of the wood- characteristics is scarce for boreal forests. peckers, and by the presence of the nest sites. The We explored and documented the main charac- centres of territory sites for the study period were teristics of the nest trees and forest types using a defined by locations of annual territories. The esti- large data set from southern Finland. As the spe- mated total number of territory sites during the cies number and composition of potential nest cav- study period was 29 with a mean of about 15 annu- ity trees greatly differ between boreal and more ally occupied territories within the whole study southern forest areas (Hågvar et al. 1980, Remm & area. Lõhmus 2011, Weso³owski & Martin 2018, Pak- Based on the annual occupancy rates of territo- kala et al. 2019), we anticipate that detailed ries and estimates of nesting success (T. Pakkala, knowledge of the characteristics of nest trees and unpublished data), the data in this study covered of their spatial variation in different types of fo- ca. 20% of all nesting attempts within the study rests and in different geographical locations is im- area during the study period. All surveys of the portant for explicit guidelines for forest manage- nest trees and cavities (see below) were carried out ment and conservation. by author TP.

2. Material and methods 2.3. Nest tree and cavity data

2.1. Study area All the trees with cavities where nesting of the Grey-headed Woodpecker was observed during The study area (170 km2) is located within the the study period were classified as nest trees. We southern boreal vegetation zone in southern Fin- included only those cavities where Grey-headed land (around 61°15’N; 25°03’E; see Pakkala et al. Woodpeckers reached at least the egg-laying 2017). It is dominated by managed coniferous fo- phase, e.g. we omitted cavities where nesting at- rests on mineral soils, with a mixture of stands of tempts were interrupted during excavation, al- different ages, and many small oligotrophic lakes. though they would have contained a seemingly Human settlements in the area are scarce. The fo- complete nest cavity. The observed nest trees and rest management in the study area aims for timber cavities were annually followed during the study production, and the prevailing harvesting method period to check their reuse by the Grey-headed is clear-cutting. Woodpecker. The species of all the detected nest trees was identified. At each nest tree location, the main fo- Pakkala et al.: Nest trees and cavities of the Grey-headed Woodpecker 91 rest type of the site was defined in the field, based 2.4. Statistical methods on the classifications of Finnish forest and peatland types (Cajander 1949, Laine et al. 2012). Forest type and condition of the nest tree between Additionally, the proportions of forest types within the respective groups were compared with good- a 700-m radius from the centre of each territory ness-of-fit tests. The distribution of DBH and the site, corresponding the approximate mean terri- height distribution of the cavity holes in various tory size during the breeding season (T. Pakkala, tree species deviated from normal distribution be- unpublished data), were calculated using land cause of skewness and/or kurtosis, and therefore, cover and forest classification data (Vuorela median-based Kruskal-Wallis or Mann-Whitney 1997), digital topographic maps of the area made tests were used in the comparisons of DBH and by the National Land Survey of Finland, aerial height of the cavity holes between groups. In post photographs and field information from the study hoc comparisons between pairs after a significant area (see Pakkala et al. 2014). result, either a Bonferroni-corrected level p <0.05 We used three classes of the condition of the in the comparisons of proportions (goodness-of-fit nest tree (see also Pakkala et al. 2018). 1) Healthy: tests) or Dunn’s test with Bonferroni correction mainly a vital tree with no signs of decay; small (Kruskal-Wallis tests) was used. Spearman’s rank wounds or damages by external factors possible. correlation coefficient was used in testing the de- 2) Decaying: tree alive, but clear signs of decay pendence between the DBH and height of the cav- visible, e.g. dead branches in the crown and/or top ity holes. All statistical analyses were performed defoliation or needle loss detected. 3) Dead: tree with IBM SPSS Statistics 25. not alive. At the tree level, we used the condition from the year when the first cavity with a nesting attempt was observed during the study period. At 3. Results the cavity level, the first nesting attempt year of each cavity was applied to describe the condition 3.1. Nest tree species and their forest types of the nest tree. The cavity type was divided to two classes A total of 76 nest trees of five deciduous tree spe- whether the cavity was in the main trunk, or in cies, 80 nest cavities, and 118 nesting attempts branch of the tree. We also assessed visually were found in the study (Table 1). Aspen (Populus whether the cavity opening was excavated to an in- tremula) was dominant (70–76%) in the tree, cav- jury spot or a healthy site in the bark layer of the ity, and nesting attempt numbers. Other tree spe- tree. Injury spots included, besides various cies were birch (Betula spp.), Scots pine (Pinus wounds or scars in the trunk of tree, also basal sylvestris), grey alder (Alnus incana), and black al- areas at the sites of broken or fallen branches. der (Alnus glutinosa), and they were clearly less Nest tree diameter at breast height (DBH, 1.3 abundant (Table 1). The Grey-headed Wood- m above the ground) in the first year of nesting was peckers mainly used the trunk of the tree as the used in the analysis. The heights of cavity holes < cavity excavation site, and only two cavities 4.0 m were measured either with a rigid measuring (2.5%) were in a large branch of the tree; both were tape, a telescopic pole or a long stick of known in aspen. length with an accuracy of 0.1 m. Heights between Grey-headed Woodpeckers used more than 4.0 m and 6.0 m were estimated by measuring the one nest cavity in three of the nest trees (4% of all 4.0 m level and then estimating the remaining nest trees) during the study period. These multi- height, with an accuracy of 0.2–0.5 m. Heights > cavity trees were all aspens; in two trees two cavi- 6.0 m were estimated by a standard stick method ties, and in one tree three cavities were used in sep- (West 2009) with an accuracy of 0.5 m; 2–3 re- arate years. The observed 80 cavities included 66 peated measurements of the same cavity from dif- new cavities, 10 cavities in which the first detected ferent directions were done to decrease the error in nesting attempt definitely was in an old cavity, and measurements. 4 cavities in which it probably was in an old cavity. Moreover, in later study years, 38 nesting attempts of Grey-headed Woodpeckers were recorded in 92 ORNIS FENNICA Vol. 97, 2020

Table 1. Number of nest tree species of the Grey-headed Woodpecker and their percentages (in parentheses) in cavity trees, cavities, and nesting attempts (nestings).

Tree species Trees Cavities Nestings

Aspen (Populus tremula) 53 (70) 57 (71) 90 (76) Birch (Betula spp.) 11 (14) 11 (14) 14 (12) Grey alder (Alnus incana) 4 (5) 4 (5) 5 (4) Black alder (Alnus glutinosa) 2 (3) 2 (2) 2 (2) Scots pine (Pinus sylvestris) 6 (8) 6 (8) 7 (6)

Total 76 (100) 80 (100) 118 (100)

Table 2. Number and percentage (in parentheses) of Grey-headed Woodpecker nest trees in different forest types. The first row of each tree species shows the percentage of forest type for each cavity tree species (summing up to 100% for tree species over forest types). The second row shows the percentage of each cavity tree species of all tree species within the respective forest type (summing up to 100% within the forest type). The percentage of forest types (last row) represent their percentages of the total area in the 29 territory sites. The four forest types: MT = moist spruce dominated forests on mineral soil; OMT = moist mixed or deciduous tree-dominated forests on mineral soil; VT = dry pine dominated forests on mineral soil; SWAMP = deciduous tree-dominated, or mixed swamp forests on peatland soil.

Nest tree species MT OMT VT SWAMP

Aspen (Populus tremula) 40 (75) 12 (23) 1 (2) 0 (0) (83) (92) (14) (0) Birch (Betula spp.) 6 (55) 1 (9) 2 (18) 2 (18) (13) (8) (29) (25) Grey alder (Alnus incana) 0 (0) 0 (0) 0 (0) 4 (100) (0) (0) (0) (50) Black alder (Alnus glutinosa) 0 (0) 0 (0) 0 (0) 2 (100) (0) (0) (0) (25) Scots pine (Pinus sylvestris) 2 (33) 0 (0) 4 (67) 0 (0) (4) (0) (57) (0)

Total (n = 76) 48 (63) 13 (17) 7 (9) 8 (11) (100) (100) (100) (100) Percentage of forest type in territory sites 63 8 18 11 previously observed cavities. Thus, the percentage The use of the MT was very similar both at the of cavity reuse was 41–44%. nest tree and the territory scale (63%). However, The 76 nest trees were found in four main types the percentages of the four forest types collectively of forests (Table 2). Forests on mineral soils in- differed from the expected percentages within the cluded 1) moist spruce-dominated forests of study area (goodness-of-fit test: ¤2 = 11.7, p = Myrtillus type (MT); 2) the more fertile, moist 0.009, df = 3). OMT was favoured in relation to its mixed forests of Oxalis-Myrtillus type (OMT), in area, but the observed percentage of nest trees was which we also included the less common class of lower than expected in VT (Table 2). moist and deciduous tree-dominated forests of the Nest tree species depended on forest type: as- Oxalis-Maianthemum type; and 3) dry pine-domi- pen was a dominant in moist forests (83–92%), but nated forest of Vaccinium type (VT). Forest in dry forests Scots pine (67%), and in peatland fo- peatlands included 4) both deciduous tree-domi- rests alders (75% of nest trees) were most common nated, and mixed swamps (SWAMP) that were nest trees (Table 2). These differences in distribu- combined to a single class. tions of nest trees were significant (¤2 = 88.0, p < Pakkala et al.: Nest trees and cavities of the Grey-headed Woodpecker 93

Table 3. Number and percentage (in parentheses) of nest tree species of the Grey-headed Woodpecker according to tree condition. The condition is the situation when the first nesting attempt during the study period was observed in the tree.

Nest tree species Tree condition

Alive, healthy Alive, decaying Dead

Aspen (Populus tremula) 11 (21) 31 (58) 11 (21) Birch (Betula spp.) 0 (0) 0 (0) 11 (100) Grey alder (Alnus incana) 0 (0) 0 (0) 4 (100) Black alder (Alnus glutinosa) 0 (0) 1 (50) 1 (50) Scots pine (Pinus sylvestris) 0 (0) 2 (33) 4 (67)

Total 11 (14) 34 (45) 31 (41)

0.001, df = 12), mainly due to proportionally rest types (¤2 = 22.0, p = 0.001, df = 6). The differ- higher amounts of aspens in moist forests and both ence was mainly due to lower proportions of dead alder species in SWAMP compared with other fo- trees in moist forest types compared with the other rest types (Bonferroni-corrected p <0.05forthese types, and it was mostly caused by the above-men- pairwise differences, Table 2). tioned difference in condition between aspen and other nest tree species. 3.2. Condition of the nest trees On average, 70% of cavities were excavated at and cavity spots a visible injury spot in the tree. This percentage significantly varied depending on the nest tree The Grey-headed Woodpeckers selected mostly condition (¤2 = 15.1, p = 0.001, df = 2); 82% in decaying (45%) or dead (41%) trees for their nests; healthy and 88% in decaying, but only 45% in only 15% of nest trees were healthy (Table 3). dead nest trees (Bonferroni-corrected p <0.05for However, the condition distributions between as- the pairwise differences between dead trees and pen, the dominant nest tree species, and all the other tree groups). other tree species combined, significantly differed 2 from each other (¤ = 29.4, p < 0.001, df =2);and 3.3. Size of the nest trees the pairwise differences in all condition classes were significant (Bonferroni-corrected p < 0.05). The mean DBH of the nest tree was 37.2 cm (me- A great majority of aspens, 79%, were alive, dian value 36.0 cm, range 25.0–64.0 cm); and it whereas 87% of the nest trees of all other species varied from 27.8 cm in grey alder to 38.6 cm in as- were dead (Table 3). Moreover, the condition of pen (Table 4). There was a significant difference in the nest tree significantly differed between the fo- DBH between the tree species (Kruskal-Wallis

Table 4. Number (N), mean, standard deviation (SD), median, minimum and maximum diameter at breast height (DBH, cm) of the Grey-headed Woodpecker nest trees in various nest tree species.

Tree species N Mean SD Median Minimum Maximum

Aspen (Populus tremula) 53 38.6 9.3 36.0 25.0 64.0 Birch (Betula spp.) 11 35.7 1.9 36.0 33.0 39.0 Grey alder (Alnus incana) 4 27.81.328.026.029.0 Black alder (Alnus glutinosa)2 28.50.728.528.029.0 Scots pine (Pinus sylvestris)6 35.83.236.530.039.0

Total 76 37.2 8.4 36.0 25.0 64.0 94 ORNIS FENNICA Vol. 97, 2020

Table 5. Number (N), mean, standard deviation (SD), median, minimum and maximum heights (m above the ground) of the Grey-headed Woodpecker cavity holes in various nest tree species.

Tree species N Mean SD Median Minimum Maximum

Aspen (Populus tremula) 57 8.1 3.1 8.5 1.8 16.0 Birch (Betula spp.) 11 7.5 1.8 6.5 6.0 11.0 Grey alder (Alnus incana) 4 5.0 0.6 5.0 4.5 5.5 Black alder (Alnus glutinosa) 2 5.8 0.4 5.8 5.5 6.0 Scots pine (Pinus sylvestris) 6 7.3 2.2 7.5 4.0 10.0

Total 80 7.8 2.9 7.3 1.8 16.0

test: H = 14.5, p = 0.006, df = 4). The DBH of grey 4. Discussion alder was significantly smaller than the DBH of aspen (Dunn’s test with Bonferroni correction: p < 4.1. Nest tree species and their forest types 0.05), but all other pairwise differences in DBH were insignificant. We detected one common and four less abundant The DBH significantly differed between the nest tree species in our study, and we found that the various forest types (Kruskal-Wallis test: H = 13.3, proportions of tree species used by the Grey- p = 0.004, df = 3). The pairwise differences be- headed Woodpecker depended on forest type. tween SWAMP and MT, and SWAMP and VT Overall, aspen was the most common species were significant (Dunn’s test with Bonferroni cor- (70%) and a dominant in moist forests (83–92%). rection: p < 0.05) with smaller median DBH in However, in dry forests Scots pine (67%), and in SWAMP, but all other pairwise differences be- peatland forests alders (75% of nest trees) were tween forest types were insignificant. The DBH most common nest trees. The importance of aspen also significantly differed between nest trees with as a nest tree in boreal areas is commonly empha- various condition (Kruskal-Wallis test: H = 6.37, p sised (e.g. von Haartman et al. 1963–72, De- = 0.04, df = 2); dead trees were smaller than decay- mentiev & Gladkov 1966, Svensson et al. 1999, ing trees (Dunn’s test with Bonferroni correction: Fetisov 2017a), although there are only few quan- p < 0.05), but all other pairwise differences be- titative studies of nest tree species within these tween condition types were insignificant. areas. However, Karhumäki (1979) and Karlin (1979) found that aspen dominated as nest tree 3.4. Heights of cavity holes species with 78% (n = 27) and 77% (n = 30) in studies in SW Finland, and Jussila (1981) with The mean and median heights of all cavity holes (n 75% (n = 21) in a study in central southern Finland. = 80) above the ground were 7.8 m and 7.3 m, re- Moreover, in a study in southern coast of Finland, spectively (range 1.8–16.0 m) (Table 5). The mean the percentage of aspen was 83% (n = 18; T. heights of cavity holes between tree species varied Pakkala, unpublished data), and 70% (n = 20) in from 5.0 m in grey alder to 8.1 m in aspen (Table another study in southern Finland just south of our 5). The median height of cavity holes did not, how- study area (T. Pakkala & J. Tiainen, unpublished ever, significantly differ between the nest tree spe- data). These five studies from Finland contained a cies (Kruskal-Wallis test: H =7.3,p = 0.12, df =4), total of 116 nest trees of five different species, but the difference was significant among the forest namely aspen (77%), birch (9%), Scots pine (6%), types (H = 8.54, p = 0.036, df = 3). There was also a grey alder (5%), and black alder (3% of nest trees), significant positive correlation between DBH and which were the same as recorded in our study. The the height of cavity holes in all nest trees percentages of nest tree species observed in these

(Spearman’s correlation: rs = 0.72, p < 0.001, df = Finnish studies were also quite similar to the re- 78). spective percentages in our study. Pakkala et al.: Nest trees and cavities of the Grey-headed Woodpecker 95

It is evident that the percentages of forest types above), spruces (Picea spp.), firs (Abies spp.), and of nest trees vary depending on the biotope distri- larches (Larix spp.) are also used in some areas bution within each study area, which has also ef- (e.g. Glutz von Blotzheim & Bauer 1980, Fetisov fects on the percentages of various nest tree spe- 2017a). In addition, nest-boxes may be used for cies. We do not have comparable information nesting (e.g. Hortling 1929, Haftorn 1971, Glutz about the biotope distributions from the areas of von Blotzheim & Bauer 1980), but these cases are the above-mentioned Finnish studies. However, in apparently quite rare. the study located next to our area, the forests are generally more fertile compared with our study 4.2. Selection of nest trees area especially with a larger proportion of OMT and more fertile forests (Soveri 1933). Thus, in Nest cavities in our study were mostly (85%) in this area, as much as 60% of nest trees located in decaying or dead trees. However, a great majority OMT (T. Pakkala & J. Tiainen, unpublished data) of aspens (79%) were living, whereas only 13% of whereas the respective proportion was only 17% the nest trees of all other species were alive. The in our study although the OMT was selected more proportions of living nest trees were higher in often than expected. fresh forest types, where aspen dominated among In other quantitative studies from boreal and nest trees. The situation was similar in the nearby hemiboreal areas of Europe, the percentage of as- study area with 93% living aspens, but only with pen of nest trees was 87–91% in Norway with 17% living other nest trees (T. Pakkala & J. three other, occasional tree species (Haftorn 1971, Tiainen, unpublished data). There are no exact Hågvar et al. 1990, Stenberg 1996), and 40–95% data of nest tree condition in other Finnish studies, in European parts of Russia with eight less abun- but majority of aspens were mentioned to be alive dant tree species (Fetisov 2017a). Local differ- in two studies (Karlin 1979, Jussila 1981). In stud- ences were also observed, e.g. in a few areas in ies with aspen as a dominant tree species the pro- Norway and in Russia, the Grey-headed Wood- portion of dead and dying trees was 78–87% in pecker was mentioned to use mainly oaks Norway, although 63–78% of the nest trees were (Quercus spp.) as nest trees (Haftorn 1971, Hågvar still living (Hågvar et al. 1990, Stenberg 1996), but et al. 1990, Fetisov 2017a). the respective amount living nest trees was only In more southern, mainly temperate European 38% in Romania (Domokos & Cristea 2014). areas, where various woodpecker species prefer Although most aspens which were used for deciduous trees for nesting, the number of nest tree nest trees in our study were living, the Grey- species is usually larger than in boreal areas (see headed Woodpeckers often (70%) selected injury Hågvar et al. 1980, Pakkala et al. 2019). Glutz von spots in the tree trunk for the cavity excavation Blotzheim & Bauer (1980) and Fetisov (2017a) site, and these spots were especially common in list some 13–20 various nest tree species of the still living trees (87%). In the aspens the cavity Grey-headed Woodpecker in central Europe and holes were often located in round, basal areas of in Russia. In temperate areas of central Europe broken branches. Similar pattern was detected in beeches (Fagus spp.) and oaks are used most often central European studies with beeches as domi- as nest trees (Conrads & Herrmann 1963, Blume nant nest tree species: almost all nest trees were 1996, Südbeck 2009), but in similar areas of east- living, but the Grey-headed Woodpeckers regu- ern Europe the nest tree choice is more variable larly used injured areas, typically vertical scars in (Dementiev & Gladkov 1966, Fetisov 2017a). As- the trunk for cavity excavation (Conrads & Herr- pen can also be a dominant nest tree species in mann 1963, Blume 1996, Südbeck 2009). some of the southern areas, e.g. in a few areas in The injured areas allow a suitable spot in the northern Germany (Brand & Südbeck 1998), in trunk to start the cavity excavation to the soft inte- Romania (Domokos & Cristea 2014), and in rior, and simultaneously the firm core with callus Ukraine (Fetisov 2017a). formation around the injured areas sustain the cav- Grey-headed Woodpeckers predominantly se- ity and protect it from rain and sunshine (see lect deciduous tree species for nesting, but conifer- Conrads & Herrmann 1963). Cavities were also ous tree species, such as pines (Pinus spp.) (see located almost always in the trunk of the tree, and 96 ORNIS FENNICA Vol. 97, 2020 only occasionally in large branches, e.g. with 2.5% between nest tree diameter and cavity height. Al- (this study), 0% (Kosiñski & Kempa 2007), 5% though the diameters of cavity trees were of the (Südbeck 2009), and 13% (Domokos & Cristea same size compared with the above-mentioned 2014), which may be caused both by the location heights of other studies, cavities were higher in our of suitable injury spots, and the selection of large study area. The large heights of cavities in our trees for cavity excavation (see below). study area may be partly explained by use of higher trees, especially living aspens (T. Pakkala, 4.3. Size of nest trees and height of cavities unpublished data). Unfortunately, Grey-headed Woodpecker’s nest tree diameters or tree heights We found a median DBH of 36 cm for nest trees, were not recorded in previous Finnish studies. but grey alder was smaller than the other four tree However, the cavity heights of the Grey-headed species. Nest trees were smaller in SWAMP, and Woodpecker were comparable with those of Black dead trees were smaller compared with decaying Woodpecker (Dryocopus martius) in aspens with- ones. The mean DBH of aspen nest trees in our in our study area, but the diameters of the aspen study, 37 cm, was relatively similar to most results nest trees of the Black Woodpeckers were still in studies with aspen as a dominant nest tree spe- larger (Pouttu 1995, T. Pakkala, unpublished cies: 35 cm (Hågvar et al. 1990; 0.5 m above data). ground), 36 cm (Stenberg 1996; DBH), 29 cm (Brandt & Südbeck 1998; DBH), and 35 cm 4.4. Nest tree and cavity reuse (Domokos & Cristea 2014; DBH). We did not measure the diameter of the trunk at cavity open- Woodpeckers commonly use same, suitable trees ing, but in Norwegian studies it was 24–26 cm for nesting in various years (see Stenberg 1996, (Hågvar et al. 1990, Stenberg 1996), and in Russia Winkler & Christie 2002, Pakkala et al. 2017). We usually 20–37 cm (Ivanchev 2005). observed that Grey-headed Woodpeckers used ca. We observed a mean height of 7.8 m for cavity 4% of nest trees for nesting in several cavities dur- holes, although the range (1.8–16 m) and variation ing the study period, and many of its nest trees in the means of individual tree species (5.0–8.1 m) were also used by the Black Woodpecker and were relatively large, with the lowest mean values Great Spotted Woodpecker (Dendrocopos major) in grey alder, and the highest observed in aspen. (Pakkala et al. 2020). However, the observed Cavities were higher in MT than in SWAMP, amounts of multicavity trees generally depend on which was dominated by alders as nest trees. In the number of study years, and, especially, how of- other studies in Finland, a mean height of 5.0 m ten older nest trees have been systematically moni- was reported by Karhumäki (1979), 4.5 m by tored. In comparison, percentages of multicavity Karlin (1979) and 4.9 m by Jussila (1981). trees in the Three-toed Woodpecker (Picoides In addition, the nest card data of the Finnish tridactylus) and Lesser Spotted Woodpecker Museum of Natural History contained information (Dendrocopos minor) were 25% and 7%, respec- of the cavity height of 91 Grey-headed Wood- tively, within the same study area and nearly same pecker’s nests predominantly from southern and study period (Pakkala et al. 2017, 2019). southwestern coasts of Finland; the mean cavity Grey-headed Woodpeckers are generally re- height was 5.3 m (range 1.2–12 m) in this dataset. ported to use old cavities for nesting (e.g. Hortling In other European studies with aspen as the domi- 1929, von Haartman et al. 1963–72, Haftorn 1971, nant nest tree, the mean cavity height was consis- Glutz von Blotzheim & Bauer 1980), but quantita- tent with the other Finnish results; 5.6–6.4 m tive data of cavity reuse are rare. We observed that (Hågvar et al. 1990, Stenberg 1996, Brandt & reuse rate of old cavities was 41–44%, which is Südbeck 1998, Ivanchev 2001, Domokos & Cris- high compared with results of other European tea 2014). In a review study from Russia, Fetisov woodpecker species (see Wiebe et al. 2006, (2017b) reported a total range of cavity height of Pakkala et al. 2017, 2020), and also within the spe- 0.5–18 m in local studies within boreal and tem- cies itself: Stenberg (1996) detected a reuse rate of perate areas in European part of Russia. 4.5% of old cavities in Norway, and Südbeck We detected a significant positive correlation (2009) that of 30% in Germany. However, Con- Pakkala et al.: Nest trees and cavities of the Grey-headed Woodpecker 97 rads & Hermann (1963) stated that most probably on an average 6–9 highly preferable aspens for the Grey-headed Woodpecker excavates a new nest trees per average territory. cavity only when old suitable cavities are not However, due to forest management history available, but they did not present any cavity reuse and habitat preferences of aspen, the observed rates. Moreover, the Grey-headed Woodpecker distribution of aspen trees is patchy within the was mentioned to excavate a new cavity more of- study area (Ahola 2005), and also elsewhere in ten than the Green Woodpecker (Picus viridis) managed boreal forests (e.g. de Chantal et al. (Glutz von Blotzheim & Bauer 1980), in which 2005, Latva-Karjanmaa 2006). There are thus Blume (1961) detected a cavity reuse rate of 89% small (or occasionally bigger) clusters of large as- in Germany. pens, and there are relatively large forest areas without any suitable large aspen trees for the Grey- headed Woodpecker (see Ahola 2005). The high 4.5. Nest trees in forest environments: reuse rate of old cavities for nesting in our study the key role of large aspens area may indicate the scarcity of good cavity trees. The lack of suitable nest trees can thus be a limit- The Grey-headed Woodpecker prefers relatively ing factor for the Grey-headed Woodpecker in large diameter, decaying or injured deciduous managed boreal forest landscapes. Conservation trees, which in managed Finnish forests are rare and retention of groups of large aspens in main compared to natural boreal forests (e.g. Nilsson et habitats, including clear-cuts, are therefore very al. 2002, Kouki et al. 2004, Vaillancourt et al. important for the continuous availability of nest 2008). In addition, our research and other above- trees. mentioned studies from boreal areas emphasise the importance of aspens as nest trees for the spe- Acknowledgements. The Lammi Biological Station pro- cies. In our study area, the density of aspens with vided excellent facilities to TP during the field seasons. DBH of at least 27 cm (and thus suitable for the The English writing was kindly revised by Dr David Wil- nest trees of the Grey-headed Woodpecker) was as son. high as 48 trees per ha in unmanaged forests, but only 0.6 trees per ha in managed forests (Ahola 2005). Similar differences in densities of large as- Harmaapäätikan pesäpuiden ominaisuudet pens between managed and unmanaged forest boreaalisissa metsissä areas were also detected in studies of aspen in eastern Finland, although the densities of large aspens Tikkoja pidetään yleisesti metsien rakenteellisen in unmanaged forest areas were clearly smaller tai lajiston monimuotoisuuden indikaattori- tai compared with those of our study area (Kouki et avainlajeina, koska ne kovertavat pesäkoloja, jot- al. 2004, Latva-Karjanmaa et al. 2007). ka ovat myöhemmin niiden itsensä ja muiden ko- According to Ahola (2005), the percentage of lolintujen käytettävissä. Kolopuiden ominaisuuk- dead aspens in aspen mappings within our study sia on kuitenkin tutkittu melko vähän. Tämä kos- area was 36% in unmanaged forests, but only 2.6% kee erityisesti boreaalisella havumetsäalueella pe- in managed forests. Moreover, in managed forests, siviä harmaapäätikkoja. 38% of aspens were classified as damaged, but Tutkimme harmaapäätikan pesäpuiden omi- most of these damages were caused by browsing naisuuksia 170 km2:n kokoisella alueella Evolla of the moose (Alces alces), which was directed Etelä-Suomessa vuosina 1987–2019. Aineisto kä- predominantly to small diameter aspens (DBH < sitti 76 pesäpuuta ja niissä 80 pesäkoloa. Harmaa- 21 cm); other types of damages were observed in päätikan havaituista 118 pesinnästä 44 % oli van- 7% of all aspens. Assuming that 10% of large as- hoissa harmaapäätikan koloissa. Pesiä löytyi nel- pens with DBH > 27 cm are damaged in managed jäntyyppisistä metsistä, mustikkatyypiltä (MT), forests, and that a typical breeding territory size of oravanmarja-mustikkatyypiltä (OMT, pinta-alaan the Grey-headed Woodpecker within the managed nähden odotettua useammin), puolukkatyypiltä forests of our study area is 1–1.5 km2 (T. Pakkala, (VT, odotettua harvemmin) ja korvista. Pesäpuu- unpublished data), we can estimate that there are lajeja oli viisi ja haapa niistä yleisin (70 % puista ja 98 ORNIS FENNICA Vol. 97, 2020

71 % koloista). Muut puulajit olivat koivu, mänty, Cajander, A. K. 1949: Forest types and their significance. harmaaleppä ja tervaleppä. Pesäpuiden lajisuhteet — Acta For. Fenn. 56: 1–71. vaihtelivat metsätyyppien kesken. Haapa oli ylei- de Chantal, M., Kuuluvainen, T., Lindberg, H. & Vanha- Majamaa, I. 2005: Early regeneration of Populus tre- sin MT:llä ja OMT:llä, mänty VT:llä ja lepät korpi- mula from seed after forest restoration with fire. — metsässä. 85 % pesäpuista oli kuolleita tai laho- Scand. J. For. Res. 20 (Suppl. 6): 33–42. avia, mutta valtaosa (79 %) kolopuiksi valituista Cockle, K. L., Martin, K. & Weso³owski, T. 2011: Wood- haavoista oli eläviä, kun taas 87 % muiden lajien peckers, decay, and the future of cavity-nesting verte- kolopuista oli kuolleita. brate communities worldwide. — Front. Ecol. Envi- Suurin osa koloista (70 %) oli koverrettu näky- ron. 9: 377–382. västi vioittuneeseen kohtaan rungolla. Pesäpuut Conrads, K. & Herrmann, A. 1963: Beobachtungen beim Grauspecht (Picus canus Gmelin) in der Brutzeit. — J. olivat suuria (keskiläpimitta rinnankorkeudelta Orn. 104: 205–247. (In German) 37,2 cm). Pesäkolot sijaitsivat keskimäärin 7,8 Cramp, S. (ed.) 1985: Handbook of the birds of Europe, the metrin korkeudella; korkeus ja rinnankorkeusläpi- Middle East and North Africa. Vol.4. — Oxford Univ. mitta korreloivat merkitsevästi keskenään. Tulok- Press, Oxford. set osoittavat suurten haapojen merkittävyyttä har- Dementiev, G. P. & Gladkov, N. A. (eds.) 1966: Birds of maapäätikan pesäpuina. Suurten haapojen ryhmi- the Soviet Union Vol. 1. — Israel Progr. for Scientific en säästäminen ja suojelu lajin pääelinympäris- Translations, Jerusalem. töissä, niiden hakkuualat mukaan lukien, on tärke- Domokos, E. & Cristea, V. 2014: Effects of managed forest structure on woodpeckers (Picidae) in the Niraj ää pesäpuiden pysyvän saatavuuden turvaamisek- valley (Romania): Woodpecker populations in mana- si. Tämä koskee erityisesti talousmetsiä, missä so- ged forests. — North-western Journal of Zoology 10: pivat pesäpuut voivat olla rajoittava tekijä harmaa- 110–117. päätikalle. Drever, M. C., Aitken, K. E. H., Norris, A. R. & Martin, K. 2008: Woodpeckers as reliable indicators of bird rich- ness, forest health, and harvest. — Biol. Cons. 141: 624–634. References Fetisov, S. A. 2017a: O gnezdovykh duplakh i evolyutsii gnezdostroyeniya u dyatlov Picidae: Rol’ fautnogo Ahola, R. 2005: Haavan esiintyminen Evon suojelualuei- drevostoya i derev’yev s “myagkoy” drevesinoy dlya den ympäristössä: Maastokartoitusten tulokset ja tule- gnezdovaniya dyatlov [The nest cavities and the evo- vaisuuden ennuste (Aspens around the conservation lution of nesting in Picidae: the role of injured and areas of Evo – mapping results and aspens future). — softwood trees for nesting]. — Russian Journal of B. Sc. Thesis, Häme University of Applied Sciences. Ornithology 26 (1499): 3867–3901. (In Russian) (In Finnish with English summary) Fetisov, S. A. 2017b: O gnezdovykh duplakh i evolyutsii Aitken, K. E. H. & Martin, K. 2007: The importance of ex- gnezdostroyeniya u dyatlov Picidae: kharakteristiki cavators in hole-nesting communities: availability and razmeshcheniya gnezdovykh dupel na derev’yakh use of natural tree holes in old mixed forests of western [The nest cavities and the evolution of nesting in Pici- Canada. — J. Ornithol. 148: 425–434. dae: characteristics of the location of cavities]. — Andersson, J., Gómez, E. D., Michon, S. & Roberge, J.-M. Russian Journal of Ornithology 26 (1527): 4835– 2018: Tree cavity densities and characteristics in ma- 4856. (In Russian) naged and unmanaged Swedish boreal forest. — Glutz von Blotzheim, U. N. & Bauer, K. M. 1980: Hand- Scand. J. For. Res. 33: 233–244. buch der Vögel Mitteleuropas. Vol 9. — Akademische Blume, D. 1961: Über die Lebensweise einiger Spechtar- Verlagsgesellschaft, Frankfurt a.M. (In German) ten (Dendrocopos major, Picus viridis, Dryocopus von Haartman, L., Hildén, O., Linkola, P., Suomalainen, P. martius). — J. Orn. (Sonderheft) 102: 1–115. (In Ger- & Tenovuo, R. 1963–72: Pohjolan linnut värikuvin. man) — Otava, Keuruu. (In Finnish) Blume, D. 1996: Schwarzspecht, Grauspecht, Grünspecht. Haftorn, S. 1971: Norges fugler. — Universitetsforlaget, Dryocopus martius, Picus canus, Picus viridis. — Die Oslo. (In Norwegian) Neue Brehm-Bücherei Bd. 300, Westarp Wissen- Hågvar, S., Hågvar, G. & Mønness, E. 1990: Nest site se- schaften, Magdeburg, 5. überarb. Auflage. (In Ger- lection in Norwegian woodpeckers. — Holarctic Ecol. man) 13: 156–165. Brandt, T. & Südbeck, P. 1998: Zur ökologischen Flexibi- Hardenbol, A. A., Pakkala, T. & Kouki, J. 2019: Persisten- lität des Grauspechts (Picus canus) – ein neuer Moor- ce of a keystone microhabitat in boreal forests: caviti- vogel am Steinhuder Meer. — Vogelkdl. Ber. Nieder- es of Eurasian Three-toed Woodpeckers (Picoides tri- sachs. 30: 1–14. (In German) dactylus). — For. Ecol. Manage. 450 (article 117530). Pakkala et al.: Nest trees and cavities of the Grey-headed Woodpecker 99

Hortling, I. 1929: Ornitologisk handbok. — J. Simelii ar- dead trees in old-growth temperate and boreal forests. vingar, Helsingfors. (In Swedish) — Forest Ecology and Management 161: 189–204. Ivanchev, V. P. 2001: K ekologii sedogo diatla v Okskom Pakkala, T. 2012: Spatial ecology of breeding birds in fo- zapovednike. [On the ecology of Grey-headed wood- rest landscapes: an indicator species approach. — pecker in Oka Nature Reserve]. — Ornithologija 29: Diss. Forestales 151. 155–161. (In Russian) Pakkala, T., Lindén, A., Tiainen, J., Tomppo, E. & Kouki, Ivanchev, V. P. 2005: Sedoi diatel Picus canus (Gmelin, J. 2014: Indicators of forest biodiversity: which bird 1788) [Grey-headed Woodpecker Picus canus (Gme- species predict high breeding bird assemblage diversi- lin, 1788)]. — In: Butev, V. T., Zubkov, N. I., Ivan- ty in boreal forests at multiple spatial scales? — Ann. chev, V. P., Koblik, E. A., Kovshar, A. F., Kotiukov, J. Zool. Fennici 51: 457–476. V., Liuleeva, D. S., Nazarov, J. N., Nechaev, V.A., Pri- Pakkala, T., Tiainen, J. & Kouki, J. 2017: The importance klonsky, S. G., Pukinsky, J. B., Rustamov, A. K., Soro- of nesting cavity and tree reuse of the Three-toed kin, A. G. & Fridman, V. S., Ptitsy Rossii i sopredel’- Woodpecker Picoides tridactylus in dynamic forest nykh regionov: Sovoobraznye, Kozodoeobraznye, landscapes. — Ann. Zool. Fennici 54: 154–175. Strizheobraznye, Raksheobraznye, Udodoobraznye, Pakkala, T., Tiainen, J., Piha, M. & Kouki, J. 2018: Nest Diatloobraznye [Birds of Russia and adjacent regions. tree characteristics of the old-growth specialist Three- Volume 6: Strigiformes, Caprimulgiformes, Apodi- toed Woodpecker Picoides tridactylus. — Ornis Fen- formes, Coraciiformes, Upupiformes, Piciformes], nica 95: 89–102. pp. 309–319. Tovarishchestvo nauchnykh izdanii Pakkala, T., Tiainen, J., Pakkala, H., Piha, M. & Kouki, J. KMK. (In Russian) 2019: Nest tree characteristics of the Lesser Spotted Jones, C. G., Lawton, J. H. & Shachak, M. 1994: Orga- Woodpecker (Dendrocopos minor) in boreal forest nisms as ecosystem engineers. — Oikos 69: 373–386. landscapes. — Ornis Fennica 96: 169–181. Jussila, E. 1981: Pikku- ja harmaapäätikan pesintäaikai- Pakkala, T., Tiainen, J., Pakkala, H., Piha, M. & Kouki, J. nen esiintyminen Päijät-Hämeessä 1970-luvulla. — 2020: Dynamics of the cavities of Grey-headed Wood- Päijät-Hämeen Linnut 12: 21–25. (In Finnish) peckers (Picus canus) reveal their long- and short-ter- Karhumäki, J. 1979: Vähälukuisten tikkojen esiintymises- m ecological roles in boreal forests. — Forest Ecology tä pesimäaikana Salon seudulla 1967–1977. — Ukuli and Management (submitted). 10 (1): 32–37. (In Finnish) Pouttu, P. 1985. Palokärjen (Dryocopus m. martius) pesi- mäbiologiasta Etelä-Hämeessä vv. 1977–1984. — Karlin, A. 1979: Harmaapäätikan esiintymisestä TLY:n Kanta-Hämeen Linnut 9: 40–50. (In Finnish) alueella 1960–1978. — Ukuli 10 (3): 31–39. (In Fin- Remm, J. & Lõhmus, A. 2011: Tree cavities in forests – nish) The broad distribution pattern of a keystone structure Kosiñski, Z. & Kempa, M. 2007: Density, distribution and for biodiversity. — For. Ecol. Manage. 262: 579–585. nest-sites of woodpeckers Picidae, in a managed fo- Roberge, J.-M., Angelstam, P.& Villard, M.-A. 2008: Spe- rest of western Poland. — Pol. J. Ecol. 55: 519–522. cialized woodpeckers and naturalness in hemiboreal Kouki, J., Arnold, K. & Martikainen, P. 2004: Long-term forests deriving quantitative targets for conservation persistence of aspen – a key host for many threatened planning. — Biol. Cons. 141: 997–1012. species – is endangered in old-growth conservation Saari, L. & Südbeck, P. 1997: Grey-headed Woodpecker areas in Finland. — Journal for Nature Conservation Picus canus. — In: Hagemeijer, E. J. M. & Blair, M. J. 12: 41–52. (eds.), The EBCC Atlas of European breeding birds: Laine, J., Vasander, H., Hotanen, J.-P., Nousiainen, H., their distribution and abundance. – T. & A. D. Poyser, Saarinen, M. & Penttilä, T. 2012: Suotyypit ja turve- London, pp. 442–443. kankaat – opas kasvupaikkojen tunnistamiseen. — Siitonen, J. & Jonsson, B. G. 2012: Other associations with Metsäkustannus, Hämeenlinna. (In Finnish) dead woody material. — In: Stokland, J. N., Siitonen, Latva-Karjanmaa, T. 2006: Reproduction and population J. & Jonsson, B. G., Biodiversity in dead wood. Cam- structure in the European aspen. — Ph. D. Thesis, bridge University Press, Cambridge, UK, pp. 58–81. University of Helsinki. Soveri, J. 1933: Lammin pitäjän kasvisto. (Die Flora des Latva-Karjanmaa, T., Penttilä, R. & Siitonen, J. 2007: The Kirchspiels Lammi in Süd-Finnland.) — Ann. Bot. demographic structure of European aspen (Populus Soc. Zool.-Bot. ‘Vanamo’4(3): 1–86. (In Finnish with tremula) populations in managed and old-growth bo- German summary) real forests in eastern Finland. — Can. J. For. Res. 37: Stenberg, I. 1996: Nest site selection of six woodpecker 1070–1081. species. — Fauna norv. Ser. C. Cinclus 19: 21–38. Mikusiñski, G., Gromadzki, M. & Chylarecki, P. 2001: Südbeck, P. 2009: Beitrag zur Höhlenökologie des Grau- Woodpeckers as indicators of forest bird diversity. — spechts Picus canus. — Osnabrücker Naturwissen- Cons. Biol.15: 208–217. schaftliche Mitteilungen 35: 263–274. (In German) Nilsson, S. G., Niklasson, M., Hedin, J., Aronsson, G., Gu- Svensson, S., Svensson, M. & Tjenberg, M. 1999: Svensk towski, J. M., Linder, P., Ljungberg, H., Mikusiñski, fågelatlas. — Vår Fågelvärld, suppl. 31, Stockholm. G. & Ranius, T. 2002: Densities of large living and (In Swedish) 100 ORNIS FENNICA Vol. 97, 2020

Vaillancourt, M.-A., Drapeau, P., Gauthier, S. & Robert, (eds.), Ecology and conservation of forest birds. Cam- M. 2008: Availability of standing trees for large cavi- bridge University Press, Cambridge, UK, pp. 79–134. ty-nesting birds in the eastern boreal forest of Quèbec, West, P. W. 2009: Tree and forest measurement. 2nd ed. — Canada. — Forest Ecology and Management 255: Springer, Berlin Heidelberg. 2272–2285. Wiebe, K. L., Koenig, W. D. & Martin, K. 2006: Evolution Vuorela, A. 1997. Satellite image based land cover and fo- of clutch size in cavity-excavating birds: the nest-site rest classification of Finland. — In: Kuittinen, R. limitation hypothesis revisited. — Naturalist 167: (ed.), Finnish-Russian seminar on remote sensing 343–353. 29.8.–1.9.1994, Geodetic Institute, Helsinki, Report Winkler, H. & Christie, D. A. 2002: Family Picidae 97(2), pp. 41–51. (Woodpeckers). — In: del Hoyo, J., Elliott, A. & Sar- Weso³owski, T. & Martin, K. 2018: Tree holes and hole- gatal, J. (eds.), Handbook of the Birds of the World. nesting birds in European and North American forests. Vol.7. Jacamars to Woodpeckers. Lynx Editions, Bar- — In: Mikusiñski, G., Roberge, J.-M., & Fuller, R. J. celona, pp. 296–555.